CHARACTERIZATION OF THE KIIRUNAVAARA IRON ORE DEPOSIT FOR MINERAL PROCESSING WITH A FOCUS ON THE HIGH SILICA ORE TYPE B2
Kari Niiranen, M.Sc.
Chair of Mineral Processing
Department Mineral Resources and Petroleum Engineering Montanuniversitaet Leoben, Austria
DOCTORAL THESIS
Affidavit:
I declare in lieu of oath, that I wrote this thesis and performed the associated research myself, using only literature cited in this volume.
Leoben, 20October 2015
Kari Niiranen, M.Sc.
ACKNOWLEDGEMENT
This thesis would not have been possible without support, supervision and encouragement from many friends and colleagues at LKAB in Sweden and at the Institute of Mineral Processing, Montanuniversitaet Leoben in Austria throughout the years. A huge amount work has been done during this study and many people have performed their best at LKAB to get new information about the Kiirunavaara iron ore deposit.
First of all, my sincerest thanks go to Dr. Andreas Böhm at the Institute of Mineral Processing, Montanuniversitaet Leoben, for your support. I’m deeply grateful for your time and interest in this study.
Your excellence in mineral processing and guidance, your encouragement and immense knowledge during the research work made it possible for me to reach the target. Working with you was a true pleasure and I hope for a good cooperation also in the future.
Furthermore, I would like to thank Professor Dr. Helmut Flachberger and the staff, my colleagues, and friends at the Institute of Mineral Processing, Montanuniversitaet Leoben, for encouragement and for a nice time working together. During this study many international contacts, especially in Austria, could be established and they provided valuable input to various aspects of this work. I have encountered a very open mind which is much appreciated.
I especially acknowledge Dr. Andreas Fredriksson for you encouragement and support during this work and for reading and valuable comments on this thesis. I also want to thank Susanne Rostmark for interest and for allowing me the time to write this thesis. I would like to thank my colleagues Therése Lindberg, Charlotte Mattsby, David Alldén, and Dr. Henrikki Rutanen for your interest and the staff at LKAB’s mineral processing laboratory in Malmberget to name but a few. Thanks also to all my colleagues at LKAB R&D who have always encouraged me during this work. Furthermore, I acknowledge Christopher Gordon for language checking and comments.
Last but not least I would like to thank Dr. Heinrich Mali at the Institute of Geology and Economic Geology, Montanuniversitaet Leoben, for interesting discussions, Sandra Haslinger for your help at the laboratory work in summer of 2010, and all my dear friends in Finland, Sweden, Austria and Germany for encouragement during these years.
KURZFASSUNG
Die Eisenerzlagerstätte am Kiirunavaara zeigt derzeit einen eher niedrigen SiO2-Gehalt (ca. 2,3 bis 3,3%
SiO2), aber es wird erwartet, dass er in den tieferen Teilen der Lagerstätte ansteigen wird. Im Jahr 2007 wurde ein Projekt „Silica in the Mine“ mit dem Ziel gestartet, ein vereinfachtes Verfahren im Labormaßstab in Bezug auf Aufbereitungseigenschaften zu entwickeln, um den Energieverbrauch und den SiO2 Gehalt der Magnetit-Konzentrate und weitere aufbereitungstechnische Eigenschaften des Roherzes vorherzusagen.
Das erste Ziel dieser Studie war es, die Zerkleinerungs- und Aufbereitungstests, die im Ramen des Projekts „Silica in the Mine“ im Aufbereitungslabor von LKAB durchgeführt wurden, mit den Zerkleinerungs- und Aufbereitungstests im Labor des Lehrstuhls für Aufbereitung und Veredlung, Montanuniversität Leoben, zu vergleichen und zu ergänzen. Für diesen Zweck wurden drei ausgewählte Proben, die die Haupterztypen (B1, B2 und D) der Lagerstätte repräsentieren, nach der Methode der
„Optimierten Zerkleinerungskette“ (OZK) zerkleinert. Daraus kann geschlossen werden, dass diese drei Erztypen einen Unterschied in ihrer Bruchcharakteristik aufweisen, was durch die Unterschiede in der Korngrößenverteilung der Zerkleinerungsprodukte sowie unterschiedlichen in dem maßspezifischen Energieverbrauch angezeigt wird. Eine wichtige Erkenntnis aus der Zerkleinerungstests im Labor des Lehrstuhls für Aufbereitung und Veredlung ist das untypische, korngrößenabhängige Zerkleinerungsverhalten. Dieses könnte mit der Textur des Magnetiterzes erklärt werden.
Da offensichtlich der Anteil des SiO2-reichen Erztyps B2 in Zukunft zunehmen wird, wurden detaillierte Aufbereitungstests wurden für die prozessmineralogische Charakterisierung dieses Erztyps im Labor des Lehrstuhls für Aufbereitung und Veredlung im Sommer 2010 durchgeführt. Diese Tests kombinierten geologische, mineralogische und geochemische Daten mit aufbereitungstechnischen Eigenschaften wie Energieverbrauch, Korngrößenverteilung, Aufschlussgrad und Verwachsungen. Ein wesentlicher Teil dieser Studie waren die mineralogischen Untersuchungen mittels automatisierter Mineralogie (QEMSCAN®), um die modale Mineralogie, die Verteilung der Silikaten in den verschiedenen Korngrößenklassen nach der Zerkleinerung, die Verteilung von Silizium (Si) zwischen verschiedenen Silikaten sowie Aufschlussgrad und Verwachsungen von Magnetit und Silikaten zu untersuchen. Ein wichtiges Ergebnis dieser Studie war die Entdeckung von zwei Subtypen im Erztyp B2, die sich durch das Auftreten von Aktinolith im Subtyp B2-a unterscheiden. Das wird als Ursache für die unterschiedliche Mahlbarkeit (d.h. Bruchcharakteristik) dieser beiden Subtypen angesehen.
SCLAGWORTE: Aufbereitung, Aufschlussgrad, Davis Rohrscheider, QEMSCAN, automatisierte Mineralogie, Magnetit, Aktinolith, LKAB, Kiirunavaara
ABSTRACT
The Kiirunavaara iron ore deposit shows a rather low content of silica (ca. 2.3 to 3.3% SiO2) but the silica grade is expected to increase in the deeper parts of the deposit. A project called “Silica in the Mine” was started in 2007 with the target to develop a simplified method in laboratory scale to predict the energy consumption and SiO2 grade in the magnetite concentrate at the industrial scale and further predict the physical properties of the crude ore with respect to mineral processing characteristics.
The first target of this study was to control and compare the comminution and mineral processing tests in relation to the “Silica in the Mine” project carried out at LKAB’s mineral processing laboratory to the comminution and mineral processing tests at the laboratory of the Institute of Mineral Processing, Montanuniversitatet Leoben. For the purpose of this study, three samples were selected representing three main ore types (B1, B2, D) of the Kiirunavaara iron ore deposit. They were first were first ground according to the “Optimized Comminution Sequence” (OCS) method. It can be concluded that these three samples, representing different ore types, show a difference in their breakage behavior based on the ore characterization data defined by the differences in the particle size distribution within comminution products, as well as in differences in the mass-specific energy consumption. It is significant to note based on the information from the comminution test that there might be a deviant breakage characteristic in relation to the iron ore from the Kiirunavaara deposit, which can be explained can be explained with the crystal structure of magnetite. Furthermore, a separation with Davis magnetic tube was a crucial part of the mineral processing test to study the liberation of magnetite and silicates.
As it was evident that the amount of the high-SiO2 ore type B2 increases, detailed mineral processing tests were carried out for process mineralogical characterization of this ore type at the laboratory of the Institute of Mineral Processing in summer of 2010. These tests combined geological, mineralogical, and geochemical information with mineralogical processing characteristics such as energy consumption, particle size distribution, and liberation and intergrowths. The essential part of this study was the mineralogical investigations using automated mineralogy (QEMSCAN®) to study the modal mineralogy, the distribution of silicates in the different particle size classes after comminution, the deportment of silicon (Si) between various silicates and degree of liberation and intergrowth of magnetite and silicates. An important result of this study was the discovery of two separate subtypes within ore type B2 based on the occur of actinolite in the subtype B2-a. This can be considered as the cause of the difference in grindability (i.e., characteristic) of these two subtypes.
KEYWORDS: mineral processing, liberation analysis, Davis magnetic tube, QEMSCAN, automated mineralogy, magnetite, actinolite, LKAB, Kiirunavaara.
ACKNOWLEDGEMENT KURZFASSUNG
ABSTRACT
1. INTRODUCTION ... 1
1.1 Background ... 1
1.2 Previous studies ... 2
1.2.1 Mineral processing model for the Leveäniemi open pit ... 2
1.2.2 Comparing the comminution at laboratory scale to pilot and large scale ... 4
2. OBJECTIVES ... 4
3. THE KIIRUNAVAARA IRON ORE DEPOSIT ... 6
3.1 Geology and mineralogy ... 6
3.2 Ore types... 8
3.2.1 Preface ... 8
3.2.2 Ore type B1 ... 8
3.2.3 Ore type B2 ... 9
3.2.5 Ore types D1, D3 and D5 ... 10
3.3 SiO2-bearing minerals ... 11
3.3.1 Preface ... 11
3.3.2 Actinolite Ca2(Mg,Fe)5SiO8O22(OH,F,Cl)2 ... 12
3.3.3 Phlogopite KMg3(Si3Al)O10(F,OH)2 ... 13
3.3.4 Chlorite (Mg,Al,Fe)12[(Si,Al)8O22](OH)16 ... 15
3.3.5 Titanite CaTiSiO5 ... 16
3.3.6 Quartz SiO2 ... 16
3.3.7 Alkali Feldspar (Albit Na[AlSi3O8] and K-Feldspar K[AlSi3O8]) ... 17
3.3.8 Talc Mg6[Si6O20](OH)4 ... 18
4. COMMINUTION TESTS ... 18
4.1 Sampling ... 18
4.2 Optimized Comminution Sequence (OCS) ... 20
4.2.1 Samples ... 20
4.2.2 Optimized Comminution Sequence (OCS) ... 20
4.2.3 Comminution tests ... 21
4.2.3.1 Laboratory rod mill... 21
4.2.3.2 Laboratory ball mill ... 23
4.3. Comminution in an open circuit ... 24
4.3.1 Samples ... 24
4.3.2 Comminution tests for characterization of ore type B2 ... 24
4.3.2.1 Laboratory rod mill... 24
4.3.2.2 Laboratory ball mill ... 25
4.4 Comminution tests related to the “Silica in the Mine” project ... 25
5. MINERAL PROCESSING TESTS ... 27
5.1 Separation with Davis magnetic tube ... 27
5.1.1 Preface ... 27
5.1.2 Davis magnetic tube tests at the laboratory of the Institute of Mineral Processing ... 29
5.1.3 Davis magnetic tube tests for characterization of ore type B2 ... 30
5.1.4 “Silica in the Mine” project ... 30
5.2 Specific gravity ... 31
5.2.1 Determination of specific gravity with He-gas Pycnometer ... 31
5.2.3 Determination of specific gravity at the laboratory of the Institute of Mineral Processing 33 5.2.4 “Silica in the Mine” project ... 33
5.3 Specific surface area ... 34
5.3.1 Preface ... 34
5.3.2 Determination of specific surface area with Blaine method and Permaran ... 34
5.3.3 Svensson’s method related to the “Silica in the Mine” project... 36
5.4 Chemical analysis ... 36
5.5 Additional mineral processing tests ... 38
5.5.1 Laser diffraction ... 38
5.5.2 Satmagan test ... 38
6. DETERMINATION OF THE DISPERSITY OF THE COMMINUTION PRODUCTS ... 40
6.1 Particle size distribution ... 40
6.1.1 Preface ... 40
6.1.2 Screen analysis ... 40
6.1.3 Comparative comminution tests at the laboratory of the Institute of Mineral Processing ... 41
6.1.4 Particle size distribution of ore type B2 ... 43
6.1.3 Particle size distribution related to “Silica in the Mine” project ... 44
6.2 Particle size distribution below 125 µm ... 47
6.3 Specific surface area, shape factor and P80 values ... 48
6.3.1 Specific surface area ... 48
6.3.1.1 Comparative tests at the laboratory of the Institute of Mineral Processing ... 48
6.3.1.2 Specific surface of samples for characterization of ore type B2 ... 49
6.3.2 Shape factor ... 50
6.3.3 P80 values ... 51
6.3.3.1 P80 values for comminution products at the laboratory of the Institute of Mineral Processing ... 51
6.3.3.2 P80 values related to characterization of ore type B2 ... 52
6.3.3.3 P80 values related to the “Silica in the Mine” project ... 53
7. ENERGY CONSUMPTION... 54
7.1 Optimized Comminution Sequence (OCS) ... 54
7.1.1 Preface ... 54
7.1.2 Energy consumption (laboratory rod mill) ... 56
7.1.3 Energy consumption (laboratory ball mill) ... 59
7.1.4 Mass specific energy consumption ... 60
7.1.5 Rittinger coefficient ... 61
7.1.6 Energy register diagrams ... 61
7.2 Energy consumption in open comminution circuit used for characterization of ore type B2 ... 63
7.2.1 Energy consumption (laboratory rod mill) ... 63
7.2.2 Energy consumption (laboratory ball mill) ... 64
7.2.4 Correlation between mass specific energy consumption and P80 ... 67
7.3 Energy consumption related to the “Silica in the Mine” project ... 69
7.3.1 Preface ... 69
7.3.2 Principle of the estimation of the energy consumption ... 70
7.3.3 Estimation of the energy consumption for selected samples ... 72
8. MINERALOGICAL CHARACTERIZATION OF ORE TYPE B2 ... 75
8.1 Automated mineralogy ... 75
8.1.1 Optical microscopy versus automated mineralogy ... 75
8.1.2. Principle of the automated systems for mineralogical characterization ... 76
8.1.3 QEMSCAN® Analysis ... 77
8.1.4 Mineralogical investigations at LKAB’s metallurgical laboratory ... 78
8.1.5 EMPA analysis on SiO2-bearing minerals ... 80
8.1.6 Assay reconciliation ... 81
8.1.7 Identification of magnetite and hematite ... 83
8.2 Modal mineralogy ... 84
8.3 Modal mineralogy of silicates ... 85
8.4 Distribution of SiO2-bearing minerals ... 86
8.5 Deportment of silicon (Si) ... 90
9. LIBERATION ANALYSIS ... 93
9.1 Preface ... 93
9.2 Intergrowths of magnetite... 94
9.2.1 intergrowths of magnetite and gangue mineral ... 94
9.2.2 Intergrowth of magnetite with silicate minerals ... 98
9.3 Liberation analysis based on the Henry-Reinhardt charts ... 100
9.3.1 Principle of the Henry-Reinhardt charts ... 100
9.3.2. Construction of the Henry-Reinhardt chart... 102
9.3.3 Henry-Reinhardt charts for samples 6382 (B1), 6365 (B2), and 6354 (D3) ... 103
9.3.4 Henry-Reinhardt charts for ore type B2 ... 107
9.4 Liberation analysis with QEMSCAN ... 110
9.5 Separation of magnetite and silicates ... 114
10. PREDICTION OF SiO2 GRADE FROM LABORATY SCALE TO INDUSTRIAL SCALE ... 117
10.1 Estimation of SiO2 at P80 = 45 µm at the laboratory scale ... 117
10.2 Comparison of SiO2 grade between laboratory scale and industrial scale ... 119
10.3 Function for estimation of SiO2 grade from the laboratory scale to industrial scale ... 120
10.4 SiO2 grade at industrial scale ... 122
11. DISCUSSION AND CONCLUSIONS ... 127 REFERENCES
LIST OF FIGURES LIST OF TABLES APPENDICES
1 1. INTRODUCTION
1.1 Background
LKAB (Luossavaara-Kiirunavaara AB) is operating an underground iron ore mine, three beneficiation plants (KA1, KA2 and KA3) and three pelletizing plants (KK2, KK3 and KK4) in Kiruna, Sweden.
Methods of mineral processing at LKAB are comprised of a sequence of two-stage comminution, WLIMS (Wet Low Intensity Magnetic Separation) and reversed apatite flotation, in which magnetic separation is regarded as the crucial part of silicate mineral separation from the ore. The aim of this process is to liberate and free the ore of phosphorous (apatite) and silica (silicate minerals) as far as possible (Adolfsson 2008, Adolfsson & Fredriksson 2011).
In the spring of 2007, the variations were observed in SiO2 grade in the crude ore. The increasing silica grade in the crude ore might be demanding, especially with respect to production of direct reduction pellets (LKAB DR Oxide Pellets, LKAB 2014), which is one of the main products of LKAB in Kiruna.
DR Pellets show an average silica grade under 0.75% SiO2 (Fe grade 67.9% and P grade 0.025%). The high-grade iron ore deposit of Kiirunavaara has today a rather low grade of silica in situ, -approximately 2.3 to 3.3% SiO2 (estimated from the geological block model, LKAB). However, the silica grade is expected to increase and the phosphorous level to decrease in incoming material to the cobbing plant in Kiruna in the future based on information from the 3D resource model generated from exploration and grade control drilling (Fig. 1).
Figure 1. SiO2 (in situ) estimated from the resource block model for the Kiirunavaara mine (LKAB). x-axis corresponds the mining level (z) in the coordinate system used in the mine.
The silicates are only SiO2-bearing minerals in the Kiirunavaara iron ore deposit. It can be assumed that the mineralogy of silicates and the SiO2 grade in the crude ore undoubtedly impact the SiO2 grade in the final product; i.e., the iron ore pellets and/or the iron ore fines. In addition, mineral processing parameters, such as the breakage characteristics, the specific energy consumption and the liberation of magnetite and silicate minerals, are essential for understanding the processes at the beneficiation plants
0 1 2 3 4 5 6
800 900 1000 1100 1200 1300 1400
SiO2[% in situ]
Depth m [mining level]
Lake Ore Main Ore Total
2 in Kiruna. This is especially pertinent to the high-silica ore type B2. The whole production chain from the mine to mill and to product is being observed in several projects during an ongoing study “Silica in the Mine”, which is an important area of the research and development at LKAB (Adolfsson 2008, Quinteiro 2008, Drugge 2009, Adolfsson & Fredriksson 2011). The ongoing “Silica in the Mine” project in the Kiirunavaara mine is an essential part of this study (Niiranen & Fredriksson 2012, Niiranen &
Böhm 2013).
As additional information can be noted that the present mining levels (Fig.1) are at 820 m and 849 m in the northern part of the ore body (known as Lake Ore, located north of Y1400 in the local coordinate system used in the mine) and on the level of 993 m and 1022 m in the middle and southern part of the ore body (known as Main Ore, located south of Y1400). The local coordinate system, used in the mine, has Y-values increasing to the south, X-values increasing to the east and the Z-values increasing to the depth from a reference level of 0 m located on the former top of the Kiirunavaara Mountain.
1.2 Previous studies
1.2.1 Mineral processing model for the Leveäniemi open pit
It should be noted, in fact, that the ideas of “mine to mill” and “minerals to product” are nothing entirely new at LKAB (Fagerberg & Ornstein 1962, Bergström & Anttila 1973, Niiranen & Fredriksson 2014).
In this context, it is worth highlighting two early projects, which have been carried out to create a connection between test results from the laboratory scale to the larger-scale beneficiation process. As an early example, from 1962, a systematic macroscopic examination of drill cores was made in order to produce a rough, preliminary forecast for the product outcome of beneficiation of the Leveäniemi iron ore deposit (Fagerberg & Ornstein 1962). The Leveäniemi mine ran from 1964 until 1983, when operations were discontinued due to an economic downturn. The Leveäniemi open pit has been reactivated as one of three new mines that are planned to open in Svappavaara in 2015 (Fig. 2).
Figure 2. Former Leveäniemi open pit after it has been emptied of water in 2014 (Photo: LKAB).
3 The background for the idea was to create a 3D model for ore recovery in the process at the beneficiation plant in Svappavaara as a part of the feasibility study on the Leveäniemi deposit. The model was planned to be based on data from systematically core logging for processing parameters and the preliminary beneficiation tests. Each drill core had previously been prepared by an inventory protocol, which included identification data, chemical assays, and section boundaries for the proposed mining pallets (Fig. 3).
Figure 3. Identification protocol on a drill core containing chemical assays of iron (Fe-analys), ore type and structure (Kärnstruktur) and section boundaries for the proposed mining pallets (Sektionslängd). Hematit = hematite; Malm = ore; Gråberg = waste (Fagerberg & Ornstein 1962).
Each core section was examined on the mineralogy and texture for the mineral processing properties, whereupon the probable product outcome for a concentrator flow sheet was estimated and recorded. The particular interest of the study was the possible variations in ore texture and/or the relationship between different ore qualities in the horizontal or vertical direction, which could lead to the changing of the product outcome from year to year. In addition, a mining block (10x 10 x 10 m3) will be placed around each drill hole per mining level and the mineral processing properties based on the information from drill cores should apply to the entire block (Fig. 4). Then, the product outcome could has been easily calculated through summation of benches and profiles, respectively, and the variation in different parts of the ore body could easily be examined. From this model, it would have conceivably been possible to estimate the average recovery and to study its variation in different parts of the deposit (Fagerberg &
Ornstein 1962). But it is obvious that more objective and analytical data is required to build a predictive model for processing parameters.
Figure 4. Schematic picture combing of the mining lay out and a mining block containing mineral processing parameters for the Leveäniemi open pit (Fagerberg & Ornstein 1962).
4 1.2.2 Comparing the comminution at laboratory scale to pilot and large scale
Bergström and Anttila (1973) carried out comminution tests in an open comminution system to compare the results of grinding with rod mill, ball mill, and pebble mill with waste pebbles. Further, their target was to determine the efficiency of the mills, the energy consumption and conversion ratio at the laboratory scale, and pilot scale and industrial scale (beneficiation plant) based on these tests. Their work included experiments in three different methods to determine the grinding capacity of mills, which was divided into three subsections: estimation according to empirically developed formula, braking test, and comminution experiment. Furthermore, their work is basis to estimate the energy consumption from the laboratory scale to the industrial scale and will be describe in detail in Chapter 7.3.
2. OBJECTIVES
The understanding the Kiirunavaara iron ore deposit from both a mineralogical and geochemical perspective, and not least, from a mineral processing perspective is increasingly important as the production in the mine is advancing toward deeper levels with higher concentrations of silica-rich ore.
This will result in new challenges and requirements for production at LKAB. The reliable, sufficient and enough detailed information about the mineralogy, modal mineralogy, geological context and texture of an ore and further the entire deposit is a fundamental key to understanding its potential amenability to the beneficiation process. Equally important are the mineral processing parameters such as particle size distribution, composition of particles, liberation analysis (liberation, middlings, locked), and flotation behavior of particles as they pass through a circuit from mine to mill and further into concentrate consisting one or more of following stages blasting, crushing, grinding and flotation (Henley 1983, Butcher 2010).
To overcome the problem with the periods of high SiO2 grade and fluctuations on SiO2 grade in the crude ore as described in Chapter 1, the “Silica in the Mine” project was started in the autumn of 2007.
Furthermore, it can be regarded as a pilot project to create and test a simplified methodology for similar projects related to exploiting of new iron ore deposits in the future. In the past, there have been several projects at LKAB, which have examined the impact of mineralogical and chemical characteristic of the crude ore, and, in particular, the effect of mineral composition on mineralogy and chemistry on magnetite concentrate after the beneficiation process. These investigations have had a focus, above all, on phosphorous and alkali grade in the magnetite concentrate, but there have also been some investigations that focused on silica (Adolfsson 1995, Adolfsson 1996, Andréasson 1997, Knights 2001, Moen 2007, Adolfsson 2008, Ståhlström 2008). However, the “Silica in the Mine” project, in which a systematic sampling from drill cores covers the entire deposit and the number of samples is significant, is the first project with a main focus on silica (SiO2).
5 The first issue of the “Silica in the Mine” project involved the characterization and identification of the problem related to SiO2 and to find out, if possible, the reason for the fluctuation of SiO2 grade and especially periods of high SiO2 grade in the crude ore. The second issue of this project was development of a comprehensive sampling methodology and a simplified methodology for mineral processing tests at the laboratory scale (Appendix 1) for predicting the physical and chemical properties of the ore (Adolfsson 1996, Adolfsson 2008, Drugge 2009, Drugge 2010, Niiranen & Böhm 2012). Further, both raw data from the tests and estimated processing parameters for characterization of the ore and different ore types will be stored in the ORACLE database to be implemented into the geological 3D model (resource model). However, the model has to be modified based on test results from process mineralogical investigations.
The first target of this study is to complete and compare the mineral processing tests, which were carried out in relation to the “Silica in the Mine” project carried out at LKAB’s mineral processing laboratory to the mineral processing tests at the laboratory of the Institute of Mineral Processing, Montanuniversitaet Leoben under the guidance of Dr Andreas Böhm. After the completed comminution tests according to the “Optimized Comminution Sequence” (OCS) method, the screen analysis and the determination of specific surface were carried out on the resulting comminution products. The net energy consumption and the specific surface area of selected particle size classes of the comminution product were measured at each stage as well as the particle size distribution of the feed and the comminution product to provide the data to construct the energy-register diagram. The results of these comparative mineral processing tests are presented in this study
Furthermore, the liberation analysis and the magnetic separation with the Davis magnetic tube were an essential part of this study. The results will be presented as Henry-Reinhardt charts, which can be regarded as a graphic combination of the mineralogical or chemical information and separability or physical parameters. The Henry-Reinhardt chart also provides information on the best possible separation result at a given physical property setting, as well as the intergrowth characteristics (liberation). In addition, the chemical analyses were also carried out on the feed and the comminution products..
As it was evident that the amount of the high-silica ore type B2 increases in the deeper part of the deposit and silicates are the most significant gangue minerals of this ore type, detailed mineral processing tests were carried out for process mineralogical characterization of the high-SiO2 ore type B2 at the laboratory of the Institute of Mineral Processing in summer of 2010. A laboratory scale methodology was developed for the systematic characterization of ore type B2 for mineral processing (Appendix 2). This methodology is combining geological (ore type), mineralogical (mineralogy of silicates, modal mineralogy), geochemical (mineral chemistry, distribution of elements), and process mineralogical (energy consumption, liberation, intergrowths, simulation of SiO2 grade in concentrate) characteristics.
The crucial part of the characterization of ore type B2 for mineral processing was the investigations on
6 the intergrowths and liberation of magnetite and silicates. For the liberation analysis, not only the magnetic separation with the Davis magnetic tube was used, but also automated mineralogy (QEMSCAN®) was used to analyze the degree of liberation and intergrowth of magnetite and silicates.
The focus of these investigation, which were carried out at LKAB’s metallurgical laboratory in Luleå, was on the modal mineralogy, the distribution of silicates in the different particle size classes after comminution, and the deportment of silicon (Si) between various silicates.
3. THE KIIRUNAVAARA IRON ORE DEPOSIT 3.1 Geology and mineralogy
Kiruna is the type area for the iron ore deposits with iron oxide (magnetite and hematite) and apatite as the main minerals first named the style of mineralization as “Kiruna type” by Geijer (1910, 1931).
Approximately 40 iron ore deposits of this type are known in Northern Sweden and individual deposits show an average grade of iron and phosphorous varying between 30 and 70% Fe, and 0.05 and 5% P respectively (Bergman et al. 2001). The Kiirunavaara deposit is the largest and the best known example of this type. The deposit is a high grade iron ore deposit consisting mainly of magnetite and apatite with an average grade of 63.8% Fe, and 0.4% P (estimated from the 3D resource model, LKAB) and with varying, but mostly small amounts of gangue minerals. The Kiruna-type iron ore deposits have geochemically been distinguished from magmatic and sedimentary types of iron ores by their generally low content of titanium (0.04–0.31% Ti) and their high content of vanadium (317–2310 ppm V) (Loberg
& Horndahl 1983). Hematite is only locally encountered and is mainly developed as a secondary product along grain boundaries and fractures in magnetite ore. Meanwhile there is a larger body of hematite- magnetite ore (martite) reported in the northern most part of the ore body containing about 2.4 M tons martite ore, which is so far without any economic value (Hansson 2001, Rutanen 2012, Wartbichler 2014).
The Kiirunavaara iron ore deposit is a sheet like body, north to south striking and approximately 4 km to 4.5 km long and 50 to 100 m thick with a maximum thickness of over 200 m in the northern part (Fig.
5). The ore body is well known down to a depth of -1365 m below the surface, but it extends at least down to the depth of -1800 m level below the surface in the northern part of the deposit.
7 Figure 5. Schematic picture of the Kiirunavaara orebody, seen from north (Picture: LKAB).
This high grade iron ore deposit was probably formed as an intrusive sill (Pehrdal 1994; Martinsson 2004) and the geochronologic data for the Kiirunavaara and the Luossavaara deposits indicates the emplacement at the period between ca. 1880 Ma and 1900 Ma (Cliff et al. 1990, Romer et al. 1994).
Younger ages of 1624 ± 39 Ma (Westhues et al. 2013) and 1638 ± 39 Ma (Aupers 2014) probably represent a secondary hydrothermal overprint.
Magnetite (Fe3O4) is as of today the only ore mineral of economic value in the Kiirunavaara deposit.
The composition of the most important minerals in the deposit, magnetite and apatite, has been studied in several occasions from the economic point of view. Apatite might be economically interesting as a possible source for phosphorous in the future (Pålsson & Fredriksson 2012), but also because it consists REEs (Frietsch & Pehrdahl 1995, Martinsson et al. 2012). A body of martite (hematite-magnetite) ore has recently been discovered in the northern part of the deposit as mentioned above but so far, the martite mineralization is not considered to be economic.
In this context, only a few previous studies have been carried out to characterize the gangue mineralogy of the Kiiirunavaara iron ore deposit in recent years (Jarousseau & Pålsson 2000, Andréasson 1997, Knights 2001). Most of these mineralogical investigations are focused on REE and apatite (Smith et al.
2009, Martinsson 2011, Pålsson & Fredriksson 2012, Martinsson et al. 2012). Besides apatite, green minerals of the amphibole group, mostly actinolite, are the most common gangue minerals described in the deposit. Phlogopite, titanite, ilmenite, rutile, quartz, talc, albite and Ca-sulphates (mostly anhydrite and occasionally gypsum) may occur, but commonly in lesser quantities as well as carbonates (calcite, Fe-dolomite and ankerite) and sulfide minerals (mostly pyrite and chalcopyrite) (Niiranen 2012 b, Nordstrand 2012, Aupers 2014).
8 3.2 Ore types
3.2.1 Preface
The apatite-magnetite ore in the Kiirunavaara deposit is divided into two main types (low phosphorous ore and high phosphorous ore) from the practical and also the historical point of view (Geijer 1931).
The low phosphorous ore is further divided into two subtypes: ore type B1 (low phosphorous, low silica,) and ore type B2 (low phosphorous, high silica). The high phosphorous ore, ore type D, is divided into three different subtypes based on phosphorous grade (Table 1).
Table 1. Limit values, % Fe and % P, for different ore types of the Kiirunavaara deposit (Niiranen 2006, Niiranen
& Fredriksson 2012) and mean values for SiO2 estimated from the Oracle database (LKAB).
ORE TYPE Fe% P% X SiO2%
B1 (low P, low SiO2) > 66 < 0.1 1.8 B2 (low P, high SiO2) > 50 < 0.1 5.6
D1 (high P) > 50 0.1 – 0.8 4.3
D3 (high P) > 50 0.8 – 2.2 1.2
D5 (high P) > 50 > 2.2 1.0
However, SiO2 grade is not taken into account in this classification. Also displayed in the Table 1, the mean values for SiO2 grade in the different ore types are estimated from the assays on drill cores stored in the Oracle database by LKAB. These limit values are used when the drill core data with analysed sections is visualized in MicroStation using a software (GeoCad) developed by Propak AB. But by the reporting of drill cores from exploration and grade control drilling and mapping of geology underground, these different ore types are used based on the macroscopic mineralogy of the ore. The difference between ore types based on the mineralogy will also be described in more detail in this chapter.
Until 2009, the these different ore types were mined separately in the Kiirunavaara mine using a mining method called large-scale sublevel caving (Wimmer & Niiranen 2005, Wimmer 2012, Niiranen 2012 a). However, because of the increasing production of the crude ore from 22.3 M tons (in 2000) to 28.4 M tons (in 2014), different ore types are now mixed together by mining and only one type of crude ore is hauled (Niiranen 2012 a).
3.2.2 Ore type B1
The typical appearance of this ore type is massive, dark greyish, and very homogeneous, most often containing ca. 95 area-% of magnetite. Because of the high grade of magnetite, the density is also high (ca. 5 g/cm3). The grain size of magnetite seems to be significantly less than 1 mm. Usually, gangue minerals show a grain size larger than the fine-grained magnetite (Fig. 6). The most common gangue mineral of the ore type B1 that closely associated with magnetite is mica (phlogopite), which sometimes shows retrograde reaction to chlorite (Nordstrand 2012, Aupers 2014). Other gangue minerals associated with magnetite are titanite, quartz, minerals of the amphibole group (mostly actinolite), ilmenite, rutile,
9 carbonates and, locally, Ca-sulphates. Hematite is a minor component and occurs most often as needle- shaped crystals forming in fracture-related veinlets (Nordstrand 2012, Aupers 2014).
Figure 6. Ore type B1 (drill core ø ca. 29 mm). (A) Sample 6127: some veinlets of gangue minerals (actinolite, talc) cutting the ore; (B) Sample 6524: some very fine cracks filled by carbonate minerals cutting the ore; (C) Sample 6139: fine-grained calcite, phlogopite, titanite and sulphides between magnetite crystals (Photo: K.
Aupers).
Actinolite is typically coarse-grained (> 500 μm), and phlogopite , talc and quartz are closely associated with it (Fig. 6 A). The intergrowth of titanite and magnetite is characteristic for this ore type, and titanite often grows along the edges of magnetite. Another Ti-bearing mineral typical for this ore type is ilmenite which often occurs as fine inclusions in magnetite (Niiranen 2012 a, Nordstrand 2012, Aupers 2014).
3.2.3 Ore type B2
This problem with silicates in the crude ore and also with high SiO2 grade appears to be linked, above all, to ore type B2 (high-silica ore) (Niiranen 2012 a, Niiranen & Böhm 2013, Aupers 2014). The ore type B2 is characterized by magnetite associated with green-coloured amphibole minerals, mostly actinolite (Fig. 7). The appearance of the ore type B2 is more heterogeneous, and its density is generally lower compared to the ore type B1. This is highly dependent on the amount and distribution of silicate minerals. Locally, gangue minerals have a grain size of ca. 1 mm, while magnetite does not differ in grain size compared to the ore type B1 (Niiranen 2012 a, Aupers 2014).
The most significant SiO2-bearing minerals in the ore type B2 are actinolite, phlogopite, chlorite, titanite and in some cases also quartz. In some cases, talc and feldspar, mostly albite but also K-feldspar, can be of importance. Zircon, allanite and thorite, which were identified in only a few cases, are uncommon.
In particular, actinolite, phlogopite and titanite, and in some cases also quartz, chlorite and albite are of importance because they are the main sources of SiO2 in the crude ore. Besides ilmenite, titanite is also an important source of TiO2, especially in the ore type B2 (Niiranen 2012 b, Niiranen 2014, Aupers 2014). The mineralogy of ore type B2 will be described in detail in Chapter 8 based on information obtained by automated mineralogy.
A B C
10
Figure 7. Ore type B2 (drill core ø ca. 29 mm). (A) Sample 6200: coarse-grained, green actinolite brecciates magnetite. (B) Sample 6196: euhedral actinolite as finely disseminated in the magnetite ore. (C) 6245: stockwork- like appearance of silicate minerals, mostly green actinolite (Photo: K. Aupers).
3.2.5 Ore types D1, D3 and D5
High phoshorous ore type D is divided into three subtypes, D1, D3, and D5 (Fig. 8), based on their phosphorous grade (Table 1), of which subtype D3 can be regarded as the most common. In the deeper parts of the ore body, ore type D occurs only in the northern-most part (Lake Ore), and the amount of this ore type decrases as the depth increases (Niiranen & Fredriksson 2012, Niiranen & Böhm 2012).
Figure 8. Ore type D (drill core ø ca. 29 mm). (A) Sample 6287: subtype D1, fine veinlets of apatite associated with calcite in magnetite, creating a network-like structure. (B) Sample 6453: subtype D3, high amounts of apatite associated with magnetite. (C) Sample 6138: subtype D5, “schlieren”-like structure of apatite-rich layers and greenish minerals (actinolite?) described by Geijer (1910) (Photo: K. Aupers).
The ore type D usually contains large amount of apatite with a grain size most often similar to magnetite.
Structures and textures (e.g., brecciated magnetite, gangue mineralogy) observed in the ore type D, are locally similar to those in the low phosphorous ore types B1 and B2. Apatite is a characteristic gangue, with varying amounts dependent on the subtype. For example, apatite content varies between 12 and 20 wt.% in the samples presenting the subtype D3. Apatite is often euhedral and grain size varies between
A B C
A B C
11 a tenth to a hundredth microns. “Ghost structure” is one very characteristic structure in ore type D in the Kiirunavaara deposit. It is defined as thin, whitish bands that consist of a tight intergrowth of magnetite and apatite (Aupers 2014). Besides apatite, monazite is other P-bearing mineral in this ore type (Martinsson 2011, Pålsson & Fredriksson 2012, Martinsson et al. 2012). Gangue minerals like amphibole, phlogopite, talc and carbonates are also common in various amounts. In general, the mineralogy is similar to the subtypes of ore type D.
3.3 SiO2-bearing minerals 3.3.1 Preface
Recently, even the silicates in the Kiirunavaara iron ore deposit have been a target of mineralogical investigations (Niiranen 2012 b, Nordstrand 2012, Aupers 2014). Not least because of their increasing importance for mineral processing, when the SiO2 grade in situ increases in the deeper part of the deposit, which has a direct connection to the amount of silicates in the crude ore (Adolfsson 2008, Adolfsson &
Fredriksson 2011, Niiranen & Böhm 2013).
For this study 24 polished thin sections were produced from the provided drill core samples (6252, 6351, 6363, 6367, 6370, 6387) at the laboratory of GeoPräp in Austria. The polished thin sections were investigated with a regular petrographic polarizing microscope (Nikon Eclipse E600) using both reflected and transmitted light sources, in both unpolarized and polarized states at LKAB’s mineral processing laboratory. Special attention was paid on the identification of different minerals and especially the identification of silicates. As support for mineral identification several volumes of Rock Forming Minerals by Deer, Zussman & Howie (edit.) were used.
The most significant SiO2-bearing minerals, especially in ore type B2, are actinolite, phlogopite, chlorite, titanite and quartz. Besides ilmenite, titanite is also an important source of TiO2, especially in ore type B2 (Niiranen 2014, Aupers 2014). In some cases, talc and feldspar, mostly albite but also K- feldspar, can be of importance. Zircon, allanite and thorite, which were identified in only a few cases, are uncommon. Besides the minerals mentioned above, some clay minerals, andradite (Fe-Garnet) and stilbite have been described, but they occur very rarely. In this chapter the mineralogy of the essential silicate minerals and quartz will be described in detail. In particular, actinolite, phlogopite and titanite will be looked on in detail, because they are the main source of SiO2 in the crude ore, in some cases also quartz, chlorite and albite are of significance. Furthermore, talc can also be an important source of SiO2
in the crude ore. Other minerals containing SiO2 such as potassium feldspar, allanite, thorite and zircon, are uncommon and occur very sporadically.
12 3.3.2 Actinolite Ca2(Mg,Fe)5SiO8O22(OH,F,Cl)2
The minerals of the amphibole group are by far the most abundant silicate minerals in the Kiirunavaara deposit (Geijer 1910, Knights 2007, Niiranen 2012 b, Nordstrand 2012, Aupers 2014) and characteristic for ore type B2 (low-P, high-SiO2 ore), but they can be found in a wide range of mineral associations in different ore types (Nordstrand 2012, Aupers 2014). However, the minerals of the amphibole group seem to be less abundant in the most iron-rich parts of the deposit (Aupers 2014).
Actinolite can occur at least in two different textural forms. A part of it occurs as large coarse crystals which are partly euhedral, partly subhedral. They are often arranged in a flow-like patterns in brecciated magnetite ore (Fig. 9 A), but it can also occur as large needle-like crystals in the magnetite matrix (Fig.
9 B). The second type is probably a pseudomorph of pyroxene (clinopyroxene) resulting from metamorphism and alteration of the ore (Deer et al. 1997). Most actinolite of the latter type contains magnetite as fine inclusions (Fig. 4 B), which is expected to be an important texture when considering the liberation of magnetite and the magnetic separation with LIMS (Low Intensity Magnetic Separation) at the beneficiation plants in Kiruna. It can also be noted that very fine-grained titanite can sometimes be found at the edges of actinolite crystals at contacts with magnetite.
Figure 9. (A) Coarse-grained, subhedral, almost colorless actinolite in brecciated magnetite ore (Sample 6387.3;
transmitted light); (B) large needle-like subhedral/euhedral actinolite crystals with fine-grained magnetite inclusions (Sample 6252.4, reflected light).
A B
13 Figure 10. Chemical classification scheme of amphibole group minerals (after Leake et al. 1997). All the analyzed samples fall into the actinolite group (Aupers 2014). The green dotted line marks the EPMA analyses carried out by Nordstrand (2012).
According to the chemical classification by Leake et al. (1997) based on the Mg-number and Si (atoms per formula unit), most of the amphibole group minerals in the Kiirunavaara deposit (Fig. 10) falls within the boundaries of actinolite (Nordstrand 2012, Aupers 2014). According to Nordstrand (2012) and Aupers (2014), the rims of actinolite crystals often show a slightly elevated content of Mg compared to the core of the crystals. Si content in actinolite also varies over a wide range, especially in ore type B2 relative to actinolite in ore type D5. In some cases, a core of actinolite grains/crystals can also contain secondary gypsum and/or mica (Nordstrand 2012).
3.3.3 Phlogopite KMg3(Si3Al)O10(F,OH)2
The minerals of the biotite group (mica) are the second most abundant silicate minerals in the Kiirunavaara deposit and can be classified as phlogopite. Phlogopite belongs to the class of tri- octahedral micas with six (octahedral) ions, in which Ti can substitute for Al (Fleet 2003). Phlogopite can be present in a variety of textural positions within all parts of the deposit. The crystals are most commonly subhedral to euhedral crystals or aggregates. They are most often colorless (Fig. 11 A) or dark brown (Fig. 11 B) (Knights 2001, Niiranen 2012 b, Nordstrand 2012, Aupers 2014). Phlogopite occurs often enclosed within the magnetite, as bundles of grains in contact with magnetite breccia fragments, and in some cases, as parallel-oriented flow-like textures, very similar to those of actinolite.
Deformation features in phlogopite can be locally observed. Phlogopite sometimes displays alteration towards chlorite, which is usually one of the common alteration products of micas (Fleet 2003).
14
Figure 11. (A) Coarse-grained, colorless Mg-rich phlogopite (Sample B2-6172; transmitted light; Photo: K.
Aupers); (B) Dark brown Fe-rich phlogopite (transmitted light; Photo: J. Nordstrand).
According to Nordstrand (2012) phlogopite might also be an important source of potassium (K) in the crude ore. The content of K seems to be constant at 9.91 wt.% throughout the samples analyzed by Aupers (2014). According to Aupers (2014), Ti content varies significantly between different samples (0.08 wt.% to 1.54 wt.%) and there seems to be a negative correlation in concentration of SiO2 and TiO2
(Fig. 12) depending on the ore type. However, this trend seems to be limited mainly to the low- phosphorous ore types (B1 and B2).
Figure 12. Oxide concentration (SiO2 and TiO2) in phlogopite in different ore types based on EPMA analysis (Aupers 2014). Oxide concentrations in ore types B1 and B2 (squares) show variations in TiO2 concentrations within a sample, while D-type ores (crosses) display constant TiO2 values.
A negative correlation between SiO2 and TiO2 in phlogopite might be an important characteristics to take into consideration in the next few years, when a new or modified beneficiation process will be designed at LKAB in Kiruna and which will take into account the observation both SiO2 and TiO2 grades are increasing rapidly while the P grade is decreasing in the deeper part of the ore body. This applies in particular to flotation, which currently consists of only reverse apatite flotation. Furthermore, fluorine (F) content in phlogopite seems be constant within one sample but varies between different samples
A B
15 from 0.70–3.99 wt.%, but it incorporate not more than 0.04 atoms of chlorine per formula unit (Nordstrand 2012, Aupers 2014). Outside of apatite phlogopite can also be an important source of Cl and F in the beneficiation process.
3.3.4 Chlorite (Mg,Al,Fe)12[(Si,Al)8O22](OH)16
Chlorites are a group of phyllosilicates with a general formula, which can be summarized as A5-6T4Z18, where A = Al, Fe2+, Fe3+, Li, Mg, Mn or Ni, while T = Al or Si or a combination of them and Z = O and/or OH (Deer et al. 2009). Chlorite is generally a less common silicate compared to actinolite and phlogopite. However, in sample 6351, chlorite is the most common SiO2-bearing mineral. Most of chlorite seems to be an alteration product of phlogopite and it often occurs as alteration lamellae within phlogopite grains (Fig. 13 A). According to Aupers (2014) most chlorites in ore type B2 have a composition of brunsvigite (Fe2+,Mg,Al)6(Si,Al)4O10(OH)8. Both chlorite and phlogopite are not uncommon minerals in the ore, but they are very seldom reported macroscopically by core logging, because they normally are very fine-grained. The possible impact of chlorite on SiO2 grade in the crude ore and magnetite concentrate will be discussed in Chapter 8 in the same context as other minerals showing the same sheet-like structure, such as phlogopite and talc.
Figure 13. Chlorite as an alteration product of phlogopite in transmitted light. ( Photo: K. Aupers (A) and J.
Nordstrand (B)).
Compositionally, chlorite shows a wide range in Fe, Mg and Al content (Nordstrand 2012). The average element concentration in chlorite for Fe, Mg and Al is12.2 wt.% (FeO), 25.4 wt.% (MgO) and 18.1 wt.%
(Al2O3) respectively (Aupers 2014). In Figure 13 B, tiny needles of rutile (TiO2), called sagenite, can been seen in the chlorite. The presence of sagenite especially in chlorite is apparently related to the alteration of phlogopite to chlorite. Phlogopite can contain more than 15 wt.% TiO2. Chlorite, however, normally contains less than 0.8 wt.% TiO2 (Fleet 2003). During the alteration of phlogopite to chlorite most of Ti precipitates and crystallizes as fine rutile needles and grains (Höfig 2014).
A B
16 3.3.5 Titanite CaTiSiO5
Titanite (sphene) seems to be a common mineral in small quantities, although being fine-grained it is seldom reported macroscopically during core logging. Mineralogical studies indicate that two different generations of titanite occur in the ore (Niiranen 2012 b, Nordstrand 2012). This is the same conclusion was made by Smith et al. (2009) and Storey et al. (2007), based on data from samples of the associated rocks of the porphyry group in the Kiruna area, which are same units both in the foot and hanging wall of the Kiirunavaara deposit.
Titanite occurs mainly as bundles of fine-grained crystals along the edges of magnetite crystals and grains, but it also shows a tendency to occur as disseminated interstitial grains (Fig. 14 A) in granular magnetite (Niiranen 2012 b). This type is normally Fe-rich and Ti-poor (Knights 2001) and may represent the first (older) generation of titanite. It should be noted that part of this fine-grained mineral might be ilmenite or rutile. Titanite also occurs as coarse-grained subhedral or euhedral crystals as the second generation (Fig. 14 B) in relation to the alteration. Titanite containing the least amount of Fe, is transparent in color, while those that containing the highest proportion of Fe, usually occur at some distance from the magnetite and has a more reddish color (Nordstrand 2012).
Figure 14. (A) Fine-grained titanite as interstitial to granular magnetite and between magnetite and silicates (reflected light; see blue arrows); (B) Coarse-grained titanite as subhedral to euhedral crystals (transmitted light).
3.3.6 Quartz SiO2
Quartz seems to be a relatively uncommon mineral based on these samples. It occurs in a somewhat larger amount only in sample 6370. According to the modal mineralogy, the quartz grade falls between 0 and 0.29 wt.% (except sample 6370). It usually occurs as anhedral grains along with carbonates in narrow veins and veinlets and is typically fresh and unaltered (Fig. 15 A). Quartz in sample 6370.2 shows a distinct undulose extinction (Fig. 15 B), which can be regarded as evidence of a moderate grade of metamorphism (Deer et al. 2001).
A B
17
Figure 15. (A) Quartz grains (Qz) with carbonate (Crb) in sample 6370.2 (transmitted light); (B) Undulose extinction of quartz grains (transmitted light; polarized state).
3.3.7 Alkali Feldspar (Albit Na[AlSi3O8] and K-Feldspar K[AlSi3O8])
There are two phases of alkali feldspars described in the Kiirunvaara deposit: K-feldspar (orthoclase) and Na-feldspar (albite) (Jarousseau & Pålsson 2000, Knights 2001, Niiranen 2012 b, Aupers 2014) (Fig. 16). In this study, only some uncertain grains of K-feldspar were identified in one thin section by optical mineralogy.
Figure 16. Albite (light brown) and K-Feldspar (red) in a particle together with phlogopite and quartz; (A) BSE image; (B) Processed with a secondary SIP List presented in Figure 58 B. Sample 6370, Fraction 1.0/0.5 mm.
Albite seems to occur more frequently among subtypes of ore type B2 according to QEMSCAN analysis (Chapter 8). K-feldspar occurs most frequently with albite and phlogopite (Fig. 16 B). Alkali feldspars is commonly found in the brecciated ore at the contacts of the orebody and can end up in the beneficiation process. This type of ore is not so common except in a zone in the middle and the contact zones of the ore body. However, alkali feldspars are common constituents of the acidic and alkaline plutonic and volcanic rocks (Deer et al. 2001). Both the rocks in the foot wall (volcanic rocks of trachytic and trachy-andesitic composition) and in the hanging wall (volcanic rocks of rhyolitic and rhyodacitic
A B
A B
18 composition) are dominated by alkali feldspars (microcline and albite) (Ekström & Ekström 1997, Bergman et al. 2001).
3.3.8 Talc Mg6[Si6O20](OH)4
Talc is may also have an impact on the SiO2 grade in the magnetite concentrate at the beneficiation plants. However, talc is not as common as actinolite, phlogopite or even titanite. It is a metamorphic mineral and its occurrence is largely dependent on the availability of sufficient magnesium (Mg) (Deer et al. 2009). Only minor amounts of talc were encountered in these samples, even though it might be a more common mineral in the ore according to core logging reports.
Talc shows a small variation in chemical composition with silicon (Si) and magnesium (Mg) and a fairly high amount of Fe (3.65–5.22 wt.% FeO, respectively, 2.74–4.89 wt.%) (Nordstrand 2012, Aupers 2014). Talc occurs mainly with actinolite (Fig. 17 A) and in pores and narrow veinlets in the magnetite matrix along with carbonate minerals and phlogopite (Fig. 17 B).
Figure 17. (A) Fine-grained talc (Talc) with colorless actinolite between magnetite (Mag) grains (transmitted light); (B) massive, fine-grained talc (Talc) as an aggregate with phlogopite (Phl) (transmitted light).
Together with talc, Nordstrand (2012) has also analyzed probable clay minerals showing a composition in the range of 48–59 wt.% SiO2, 11–15 wt.% Al2O3 and 9–12 wt.% FeO which are dominant elements.
4. COMMINUTION TESTS 4.1 Sampling
The samples selected for the mineral processing tests at LKAB’s mineral processing laboratory related to the “Silica in the Mine” project were derived from drill cores from the exploration drilling and grade control drilling underground in the Kiirunavaara mine. Besides the samples from the drill cores, there was also sampling (grab samples) during the development of the tunnels at the first stage of the project.
This sampling included about 550 samples and started in the autumn of 2007 and ended at the end of
A B
19 2009. The ore type of grab samples was defined based on the information from the geological maps of the mining blocks. In general, through the sampling, it was important to keep the different ore types separate, because it was expected that the physical properties and thus also process mineralogical characteristics are unequal for different ore types. The total amount of the samples from drill cores and tunneling is about 2950. For each sample, the coordinates (x, y, z in LKAB’s coordinate system in the mine) were stored in the database to obtain spatial distribution in the ore body. Of this total amount of samples, 1815 samples are currently tested either in whole or in part to a minor extent until now.Each sample, aimed for mineral processing tests, was compounded of several subsamples after the crossing the drill cores (Ø approximately 29 mm) for chemical analysis according to the recommendation of the geologist based on the ore types after reporting drill cores.
Of these samples three samples, 6382 (B1), 6365 (B2), and 6354 (D3) representing the main ore types of the Kiirunavaara (Chapter 3.2) deposit, were selected as feed material for the comparative mineral processing tests at the laboratory of the Institute of Mineral Processing, Montanuniversitaet Leoben, based on the information from drill core logging and chemical analysis. Furthermore, six samples (6252, 6351, 6363, 6367, 6370, 6387) representing ore type B2 were selected as feed material for the process mineralogical tests at the laboratory of the Institute of Mineral Processing, Montanuniversitaet Leoben.
They originated from the middle and southern part of the ore body and were composed of several sub- samples according to the recommendations of a geologist. There were two essential arguments for the selection of these samples for this study. First, they were classified by a geologist as ore type B2 in the drill core characterized by the green based on the green actinolite in the mineral association. The second important argument, was the relatively high SiO2 grade in the crude ore and the strongly varying recovery of silica by the mineral processing tests carried out during the “Silica in the Mine” project at LKAB’s mineral processing laboratory in Malmberget. For this study, further 18 samples representing different main ore types were selected for detailed investigations to illustrate the mineral processing test in relation to the “Silica in the Mine” project and evaluate its result. Smaller groups such as martite (hematite-magnetite) or ore breccia, which are represented by only some of the samples, are excluded from the focus of this investigation.
It should be noted that in many current mineral processing test practices, the predetermined geological, mineralogical, and/or grade characteristic is used as a control for sample selection and distribution of mineral processing indices. This was also the case for sampling in relation to the “Silica in the Mine”
project. However, in several published cases, it has been demonstrated that this might be problematic.
Using predefined boundaries or limits such as lithology and/or mineralogy and grade and/or cut off of the minerals and elements of interest to control the distribution of mineral processing performances indices and test work can represent uncertainty, specially without proper evaluation of the relationship between the definition of the ore types or domains and the mineral processing characteristic (Keeney &
Walters 2011, Kittler et al. 2011). Walters (2009) has pointed out that there might be no guaranteed
20 direct relationship between the geological or mineralogical ore definition/characterization and the mineral processing performance.
4.2 Optimized Comminution Sequence (OCS) 4.2.1 Samples
The samples selected for the mineral processing tests at the laboratory of the Institute of Mineral Processing, were first crushed down to –3 mm in two stages using a jaw crusher and the MK 25 cone crusher in closed circuit at LKAB’s mineral processing laboratory. The particle size distribution for the feed material is also shown in Figure 18 A, plotted on a Gates Gaudin Schuhmann grid (GGS grid), and in Figure 18 B on a half-logarithmic grid used at LKAB.
Figure 18. (A) Particle size distribution (PSD) for samples 6382 (B1), 6365 (B2), and 6354(D3) after crushing -3 mm at LKAB’s mineral processing laboratory. (B) The same particle size distribution presented in Figure 18 A, but in the format of internal reports used at LKAB (cumulative mass, % passing is non-logarithmic). Numeric data available on CD.
4.2.2 Optimized Comminution Sequence (OCS)
A test procedure, named the “Optimized Comminution Sequence” (OCS), has been developed by Steiner (1990, 1996, 1998) at the Institute of Mineral Processing at Montanuniversitaet Leoben. The design of OCS calls for a characterization of the comminution properties of minerals and rocks with respect to their individual breakage characteristics and the specific energy consumption as a function of the product dispersity. The latter is described by the particle size distribution and the specific surface area of the products. The purpose OCS is to determine the “Natural (i.e., material inherent) breakage characteristics” (NBC) of brittle mineral matter. NBC can be defined as a size distribution with the lowest amount of fines at a given maximum particle size. Material ground at most energy efficient way by compressive stress and impact stress reveals the specific surface area/energy relationship,
1 10 100
0.01 0.1 1 10
Cumulative mass[% passing]
Particle size [mm]
6382 (B1) 6365 (B2) 6354 (D3)
0 10 20 30 40 50 60 70 80 90 100
0.01 0.1 1 10
Cumulative mass [% passing]
Particle size [mm]
6382 (B1) 6365 (B2) 6354 (D3)
A B
21 culminating in the Rittinger coefficient (R), which relates the creation of new surface during the comminution to the net energy consumption (Steiner 1991, Boehm et al. 2002).
OCS obeys the principle of the energy-optimized comminution and consists of a succession of numerous comminution stages in closed circuit designed to guarantee a small size-reduction ratio. The settings of the apparatus of each stage are optimally adapted to the specific size-reduction step. The circuit design of “closed circuit with pre-screening” (Fig. 19) is estimated at the laboratory by cyclic comminution tests at high circulating load (at least 100%). High circulating load causes a short retention time of the particles within the comminution tool, resulting in a smaller number of stress events per particle and cycle. Pre-screening serves to separate the existing fines in the feed, thus directing the energy supplied by the comminution tool to the coarse particles. Each comminution cycle ends with intermediate classification at a defined screen aperture. The accurate intermediate screening after each comminution cycle removes the fine particles soon after their creation. Within each circuit, mechanical screening is completed by manual screening, which is still the most accurate laboratory method of particle-size separation in the particle size range from 0.04 to 100 mm (Boehm et al. 2002).
Figure 19. Scheme of the closed comminution circuit design with pre-screening (S = screening, C = Comminution tool) by Steiner (1990, 1996).
4.2.3 Comminution tests
4.2.3.1 Laboratory rod mill
The comparative comminution tests were carried out with the comminution tools belonging to the standard testing equipment at the laboratory of the Institute of Mineral Processing. The OCS was built up to analyse the three samples of the Kiirunavaara iron ore on its breakage behaviour abutted to the
“Less Fines” project (Boehm et al. 2002) based on the principles of energy-optimized comminution. It consisted of two grinding stages, a laboratory rod mill and a laboratory ball mill, with pre-screening (Appendix 3). The laboratory rod mill is driven by a roller drive and powered by a frequency converter controlled electric motor and the number of revolutions was counted by a sensor (Fig. 20). The technical data for the used comminution tools is displayed in Table 2.
C S
